• Complex data structures (matrices, lists and data frames)
• Functions in R
• Reading data from files

Questions, anybody?

Summary on the lilies exercise

Elements of a vector can be accessed not only using numbers (indices) or logical vectors. You can assign names to vectors:

person <- c("January", "Weiner", 134)
names(person) <- c("FirstName", "LastName", "Age")
person["FirstName"]
person["Age"]

samples   <- c(1, 10, 23, 42, 13)
samples_n <- length(samples)

• Both c and length are functions. They take some arguments (often many of them) and return a single object: a vector, a matrix or something else.

• You can always assign the result of the function to a variable.

• Sometimes functions return NULL, which is R for “nothing”, but which is still something you can use or assign.

In R, there are two special values: TRUE and FALSE. They can be used to create logical vectors.

sel     <- c(TRUE, TRUE, TRUE, TRUE, FALSE)
sel
!sel

Comparison operators (>, <, <=, >=, ==, !=) produce logical vectors:

samples <- c(1, 1, 2, 5, 7)
samples > 2
which(samples == 7)
which(samples != 1)

Logical vectors can be used to access elements:

persons <- c("Aphrodite", "Bacchus", "Circe", "Demeter", "Eurypides")
sel     <- c(TRUE, TRUE, TRUE, TRUE, FALSE)
persons[sel]

# we can abbreviate the TRUE and FALSE to T and F (avoid)
greek <- persons[ c(T, F, T, T, T) ]

• https://youtu.be/xmeZofFlp78

• Back to the lilies - how do you change all values which are greater than 1 to 1?

Create a vector as follows:

samples <- c(1, 10, NA, 15)

NA stands for not available (e.g., missing data)

• try adding a number to that vector
• what does length(samples) return?
• what does mean(samples) return? Why is that?
• use the named parameter na.rm=TRUE for the mean() function. Look up help (?mean) to see how it can be used. What happens now?
• what does the is.na() function return when applied to samples?
• how do you find NA values? Use is.na and which
• How do you select only these samples which are not NA?

The reason we are showing how to create a function is to show you that it is simple, and also because it will help you understand what functions are.

#' Function name
#' Function description
some_name <- function(param1, param2=2) {

## code comment
# <your code goes in here>

}

• If you have to write something more than twice, you might want to turn it into a function!
• R is a functional language:
  • almost everything is a function
  • functions usually take something and return a value; they are not supposed to use or modify anything that they have not been passed to explicitely (sometimes hard to avoid)

• vectors (string, number, integer, logic, factor)
• matrices and arrays
• lists
• data frames

Much like vectors, matrices can only hold one data type (e.g. only numeric or only character or only logical etc.).

m <- matrix(1:18, ncol=3, nrow=6)
# compare with
m <- matrix(1:18, ncol=3, nrow=6, byrow=TRUE)
dim(m)
ncol(m)
nrow(m)

matrix[row, column]

So, for example:

m[1, ] # vector which is the first row
m[, 2] # vector which is the first column
m[3, 1] # first element of the third row

We can name rows and columns of a matrix and use the names to access the rows and columns:

colnames(m) <- letters[1:ncol(m)]
rownames(m) <- LETTERS[1:nrow(m)]

m["A", "b"]  # one "cell"
m["B", ]     # one row
m[   , "b"]  # one column

• If you select a single column or a single row, you will get a vector
• If you select more than one row or column, you will get a (smaller) matrix
• If you select more rows or columns than are present, you will get a “subscript out of bonds” error
• REMEMBER: Vectors and matrices always have only one data type (string, integer etc.)

Assume you have a 48 well-plate for a drug sensitivity analysis with viability scores.

• Create a 6x8 matrix “drugSensitivity” with random numbers between 0 and 1. Use matrix and runif. These reflect your viability scores.

Before starting you experiment, you decided to leave out the border wells to avoid edge effects:

• Change the values of the border elements to NA.

The rows are treated with inhibitor 1 with increasing concentrations (control, low, medium, high). Columns 2 to 4 are treated with inhibitor 2 with increasing concentrations (control, low, high) and column 5 to 7 are treated with inhibitor 3 (same concentrations as inhibitor 2).

• Use row and column names to reflect treatments.
• What are potential problems of this approach?
• Select all wells with inhibitor 3.
• Which wells can be used as negative control?
• Select only wells with a combination of inhibitor 1 and inhibitor 2.

Caveats:

• You access elements of a list using [[, not [
• Lists may have names (set with names()), but don’t have to

Data frames are a bit like matrices, but every column can store different type of data. In this, they are more like lists (which they in fact are).

names <- c("January", "Manuela", "Bill")
lastn <- c("Weiner", "Benary", "Gates")
age   <- c(1001, NA, 65)

d <- data.frame(names=names, last_names=lastn, age=age)
class(d)
class(d[,1])
class(d[,3])

You can access the data frame elements much like the elements of a matrix.

However, since data frames are lists, the list operator ($) also works:

d$names # same as d[,1] or d[, "names"]
d$lastn
d$lastn[1]

However, note that when you select a row, you will get a data frame, not a vector. This is because each of the column can be of different type, and vectors can hold only one type of data.

Caveats:

• data frames sometimes turn strings to factors (more on that later), which may have really disastrous consequences, use stringsAsFactors=FALSE
• small typos can turn a numeric vector into a string
• factors are dangerous to work with, use them cautiously

Gory details: matrices are a basic data type. Data frames are a list.

Caveats:

• tibbles are the data frames from tidyverse

• Whatever you can do to a data frame, you can do to a tibble as well

• read_* functions return a tibble

• tibble do not have row names

• If you select a single row in a data frame, you get a smaller data frame. If you select a single column, you get a vector.

• In tibble, you always get a smaller tibble.

• https://youtu.be/eWu7kvNBpyc

• Create a 5x3 matrix with random numbers. Use matrix and rnorm.
• Turn the matrix into a data frame. Use as.data.frame for that.
• Add column and row names.
• Add a column. Each value in the column should be “A” (a string). Use the rep function for that.
• Add a column with five numbers from 0 to 1. Use the seq function for that.

rep() is used to replicate vectors.

rep(c("A", "B"), 5) 
# result:
#  [1] "A" "B" "A" "B" "A" "B" "A" "B" "A" "B"

rep(c("A", "B"), each=5)
# result
# [1] "A" "A" "A" "A" "A" "B" "B" "B" "B" "B"

• [ ] … accessing the element of a vector / matrix / list / data frame -> extraction operators

• [[ ]] … accessing the element/items of a list

• $ … accessing elements by name

• ( ) … used when calling a function to provide arguments

• {} … indicating a block, eg when defining a function